DIRECTION OF FLIGHT IN 1876. 151 



"At Dallas, at noon September 20, 1876, the air was filled with the 

 first swarm of locusts; by 5 o'clock in the afternoon none were in the 

 air. Previous to this date up to the night of the 19th the wind had 

 been south ; it changed on the 20th to the northwest, and this wind 

 brought the locusts in a swarm which mnst have been many miles long 

 and broad, and from 1,000 to 2,000 feet high, as far as the eye could see. 

 At 10 o'clock, September 21, the air was again filled as at noon of the 

 preceding day, the northwest wind still blowing, an-d the grasshoppers 

 passed on as the day before, until 4 p. m. On the 22d the wind veered 

 to the south, and the locusts flew during the day in large numbers irreg- 

 ularly about, like a swarm of bees. This continued until noon of the 

 23d, when a southwest wind bore a large number to the northwest." 



They passed across Indian Territory from the north from the middle 

 of September onward, appearing at Fort Gibson September 16 to 28. 



In Benton County, in the extreme northwest of Arkansas, swarms in 

 the autumn passed over from the northwest, but none settled. 



In Missouri, according to Mr. Eiley, the locusts entered the northwest 

 corner of the State from the north and northwest early in September, 

 and were to some extent prevented from extending farther by south 

 winds. " They entered the southwest counties from the southwest nearly 

 a month later, invading Kewton and McDonald by September 23, and 

 reaching the middle of Barry by the first of October, and Cedar by the 

 middle of this month. It is quite clear that the eastern limit of the 

 swarms which came from the north and northwest was receding west- 

 ward after they reached Northwest Missouri, and that Southwest Mis- 

 souri, Southeastern Kansas, and Northwestern Arkansas would have 

 escaped had it not been for west and southwest winds that brought 

 back insects which had reached south of these points." The locusts 

 arrived a month later than in 1874, and in this respect the invasion of 

 1876 more nearly resembled that of 1866. 



In Kansas the swarms flew in the main from the north and northwest. 

 " The insects came into the northwest part of the State late in July and 

 early in August, and were seen flying about in many directions, but 

 mainly southward, during the whole month. Early in September the 

 swarms thickened, and the wind blowing almost a gale from the west 

 on the 7th and 8th of the month, and strong from the west and north- 

 west for two or three days subsequently, the insects during that time 

 swept down in darkening clouds over the greater portion of the State 

 from the 98th meridian to beyond the 96th." (Rilej 's ninth report.) " I 

 came throu<ih Kansas from Colorado (Denver) on the 5th and 6th Sep- 

 tember. Galoptenus spretus at that time extended about 100 miles east 

 of the mountains, last of which point no trace of it was to be seen dur- 

 ing daylight on the 5th. Next morning we struck locusts in small num- 

 bers at Brookville (Saline County), 180 miles west of Kansas City; in 

 full force at Salina, 12 miles farther east ; and found the east front of 

 this line 4 miles west of Abilene, iu Dickinson County, and about 150 



